Comparison of Gene Expression Profiles in Normal Retina and Brain Using a Canine Retinal cDNA Microarray

Abstract

Abstract: : Purpose: We have developed a custom–made canine retinal cDNA slide microarray containing ∼4,500 non redundant elements from a normalized canine retinal cDNA library. To validate this array for use in gene expression analysis of retinas with inherited photoreceptor degeneration, we have compared the expression profiles of normal retinas, and frontal, occipital, temporal lobes of the brain using a pooled brain RNA as a reference tissue. Methods: We employed a common reference sample based experimental design in which replicates of normal retina and frontal, occipital, temporal lobe RNA were hybridized in combination with a reference sample of pooled brain RNA. For microarray analysis we used the 3DNA Array 50TM kit. Hybridizations signals were visualized with Cy5 and Cy3 fluorescent reporter molecules, and intensities were detected using a Genepix Scanner; the fluorescence intensities of spots were analyzed using GeneSpring version 7.0. Statistical analysis was performed using SAM with an estimated false discovery rate of 20%. Northern blot analysis was used to confirm expression patterns. Results: Preliminary results were calculated separately for each set of hybridizations, comparing the relative expression of the individual tissues used to the common reference of pooled brain. Overall, each set remained with about 3,000–3,600 clones for comparison after eliminating controls and filtering invalid data. As expected, changes in expression between the sample and the reference were much higher for retina (∼10%) than the individual brain tissues (1%) that had been used to create the pool. Furthermore, the retinal samples were clearly separated from any of the hybridizations using brain tissue in a principal component analysis. The accuracy of observed changes in expression have been confirmed by northern blot analysis using six randomly chosen genes that represented a wide range of different expression between retina and brain. Conclusions: The similar expression observed between individual brain samples and the pool used as a reference for our experimental setup confirms the robustness of our chosen reference sample. The clear clustering of all retinal samples and little variance between them allowed us to establish an expression profile for normal retina using our microarray system. We will now be able to use this microarray to identify gene expression changes found in retinas of different animal models for retinal degeneration, and perform more in depth analysis to extract gene groups relevant to disease

Keywords: gene microarray • retinal degenerations: hereditary • retina